Systematic Review
Schizophrenia is associated with increased risk of medical comorbidity, possibly including osteoporosis, which is a public health concern due to its significant social and health consequences. In this systematic review and meta-analysis, we aimed to determine whether schizophrenia is associated with bone fragility. The protocol for this review has been registered with PROSPERO (CRD42020171959). The research question and inclusion/exclusion criteria were developed and presented according to the PECO (Population, Exposure, Comparison, Outcome) framework. Schizophrenia was identified from medical records, DSM-IV/5 or the ICD. The outcomes for this review were bone fragility [i.e., bone mineral density (BMD), fracture, bone turnover markers, bone quality]. A search strategy was developed and implemented for the electronic databases. A narrative synthesis was undertaken for all included studies; the results from eligible studies reporting on BMD and fracture were pooled using a random effects model to complete a meta-analysis. The conduct of the review and reporting of results adhered to PRISMA guidelines. Our search yielded 3103 studies, of which 29 met the predetermined eligibility criteria. Thirty-seven reports from 29 studies constituted 17 studies investigating BMD, eight investigating fracture, three investigating bone quality and nine investigating bone turnover markers. The meta-analyses revealed that people with schizophrenia had lower BMD at the lumbar spine [standardised mean difference (SMD) -0.74, 95% CI -1.27, -0.20; Z = -2.71, p = 0.01] and at the femoral neck (SMD -0.78, 95% CI -1.03, -0.53; Z = -6.18, p ≤ 0.001). Also observed was a higher risk of fracture (OR 1.43, 95% CI 1.27, 1.61; Z = 5.88, p ≤ 0.001). Following adjustment for publication bias, the association between schizophrenia and femoral neck BMD (SMD -0.63, 95% CI -0.97, -0.29) and fracture (OR 1.32, 95% CI 1.28, 1.35) remained. Significantly increased risk of bone fragility was observed in people with schizophrenia. This association was independent of sex, participant number, methodological quality and year of publication.